Effectuation

Effectuation (Exploration Mode)

When to Apply

• Pre-PMF stage with unclear product direction
• Post-pivot with new market entry
• Goals are ambiguous or evolving
• No historical data to estimate probabilities (Knightian uncertainty)
• Every direction looks equally plausible

Core Framework

Three-Condition Test (Auto-Evaluate)

| Condition | Causation (YES) | Effectuation (NO) | |---|---|---| | Goal clarity | Founder can state clear, measurable goal | Goal is ambiguous or evolving | | Environmental legibility | Clear which market signals matter | Every direction looks equally plausible | | Historical data | Data exists to estimate probabilities | No precedent, Knightian uncertainty |

Mode selection:

• All 3 YES: Causation mode (Strategy Kernel, KPI Tree, Critical Chain)
• Any 1+ NO: Effectuation mode (this framework)
• Mixed: Hybrid mode (Causation structure with Effectuation flexibility)

Effectuation Cycle

In effectuation mode, DISABLE: KPI tree decomposition, buffer management, Critical Chain. Run this repeating expansion cycle instead:

1. Bird-in-Hand Audit "What resources do you have RIGHT NOW?"

• Skills (founder + team)
• Network (who do you know?)
• Capital (how much can you invest?)
• Technology (what have you already built?)
• Domain knowledge (what do you uniquely understand?)

2. Affordable Loss Ceiling (4 Dimensions)

Every experiment is checked against ALL four dimensions:

| Dimension | Question | Threshold | |---|---|---| | Financial | "How much capital can you invest and survive if it goes to zero?" | Set dollar amount | | Time | "How many months can you devote before you need income?" | Set month limit | | Reputation | "Would failure here damage your ability to raise funds or get hired?" | low/medium/high | | Psychological | "Can you handle this failing and try again?" | yes/no |

Track CUMULATIVE commitments across all active experiments. Individual experiments may each be small, but together they can exceed limits.

3. Execute Small Experiment -> Measure

Design the cheapest experiment to learn something. Not the best experiment — the cheapest one that still generates a useful signal.

4. Lemonade Principle "What surprised you this week? Can it be used?"

Unexpected outcomes are not failures — they are data. A surprising customer segment, an unexpected use case, an unplanned partnership opportunity. Capture and evaluate.

5. Crazy Quilt "Who showed interest? Who self-selected in?"

Early stakeholders who voluntarily engage (customers, partners, advisors) define the venture's direction. Don't chase — follow the self-selected.

6. Resources expand -> New Bird-in-Hand -> Repeat

Goal may mutate each cycle. This is NOT scope creep — it is effectuation working correctly.

Mode Switching Triggers

Effectuation -> Causation:

• PMF confirmed (retention curve flattening)
• Founder can state clear measurable goal
• Sufficient data exists to estimate probabilities
• Announce: "Your situation has changed. Switching from exploration mode to scaling mode."

Causation -> Effectuation:

• Major pivot
• New market entry
• All 3 conditions flip to NO
• Announce: "The conditions for structured planning are no longer met. Switching to exploration mode."

Decision Rules

1. Never apply KPI trees in effectuation mode — goals are undefined, so decomposition is premature
2. Affordable loss, not expected return — ask "what can I afford to lose?" not "what's the ROI?"
3. Goal mutation is expected — don't resist it, track it
4. Follow self-selectors — Crazy Quilt participants define direction
5. Surprises are the product — Lemonade principle is the core engine of discovery
6. Cumulative exposure tracking — individual experiments are small, totals can be large

Anti-Patterns to Detect

| Anti-Pattern | Signal | Response | |---|---|---| | Premature causation | Building KPI trees with no data | "You don't have enough data to decompose this goal. Let's run experiments first." | | Ignoring surprises | Sticking to original plan despite new signals | "The lemonade principle says unexpected results are your best data. What surprised you?" | | Over-investing in exploration | Experiments getting expensive | "Check affordable loss: cumulative exposure is [X]% of your financial ceiling." | | Analysis paralysis in uncertainty | Trying to predict the unpredictable | "In effectuation mode, the cheapest experiment beats the best analysis." | | Resisting goal mutation | Treating changed goals as failure | "Goal evolution is how effectuation works. Track the mutation — is it converging?" |

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: masterleopold
• Source: masterleopold/anty-framework
• License: MIT
• Homepage: https://4n7y.com

Install and usage instructions live in the source repository linked above.